Advice reqeusted to migrate MSR from Bare Metal to Container

[cw-kid]

Vera controllers don't support VOI.

You'd have to create a scene on the Ezlo controller with an action to send a VOI command to the voice assistant.

Then in MSR you'd have to call that Ezlo scene to run via a HTTP request.

See my Idiot's guide to Ezlo HTTP Server API on the Vera forum for how to send http commands to an Ezlo hub to initiate running a scene

That is your only and shortest route as far as I can see.

[LibraSun]

I seem to recall someone describing an Alexa plug-in for the Vera which I never tried out. Which has me wondering if it could serve the same purpose although probably not as direct or robust as the ezlo-based route you describe.

[cw-kid]

Ask @therealdb that's his plugin.

I dont use it as I have Google Home and Chromecast devices.

If his plugin can send text commands to the Alexa voice assistant via a Vera hub? To then perform some action on Alexa and thus with any connected devices?

Then maybe there is some way to initiate that also from MSR?

[therealdb]

My plugin can do anything local. Plus, it will always work

I’ve been running it since almost 3 years. Tts, Alexa routines (ie to control dnd, music, weather, alarms) and direct commands. It’s always working. Look at my signature to get it.

[LibraSun]

Will do, eagerly so. Thanks!

[CatmanV2]

It's great. I really must get the OAuth done....

C

[cw-kid]

@therealdb said in Accessing VOI from MSR:

My plugin can do anything local

When you going to do a plugin for Google Home

[therealdb]

@cw-kid google has closed all the local apis. But if you found something to use, I’ll take a look.

[cw-kid]

@therealdb Google suck sometimes. Why don't they have any local APIs.

[rafale77]

@cw-kid

@cw-kid said in Accessing VOI from MSR:

@therealdb Google suck sometimes. Why don't they have any local APIs.

Good question. This move was actually the first thing that got me to remove everything google from my house from devices to browsers, search engines, email service etc... They actually did initially offer a local API but removed it for the philips hue as an example. Their strategy is to collect as much data as possible from us, customers since we are actually not their customers as part of their business model. We are their products. They are forcing people to rely on their cloud. Alexa is much better in that sense that though it is shifting to favor cloud centered environment, they have kept their local APIs open. That being said all of these cloud based environment have some major downsides: latency, reliability, privacy. I am gradually moving to local voice processing as well with homekit and it seems to turn out being a bit of a project.

@LibraSun, what devices are you looking to control? There has to be better, more direct ways to do it than to go through the amazon cloud...

[LibraSun]

@therealdb said in Accessing VOI from MSR:

My plugin can do anything local.

I'm getting mixed messages here, though, so allow me to inquire before I proceed with installing VeraAlexa plug-in...

On GitHub, the readme states to the contrary...

"On Vera only TTS is implemented."

but later suggests:

"When jq package is installed [such as with Vera f/w 7.32], the plug-in will automatically switch to the avanced version."

It also include this proviso, which is unclear to me:

"only accounts with MFA disabled"

So, which of the following functionalities will my Vera Plus actually enjoy once I install the plug-in [under 7.31], and which will only happen after upgrading to 7.32 firmware?

• Text-to-Speech (TTS) over Alexa/Echo
• Announcements to Groups
• Text-to-Command ("VOI") from Vera

(SORRY IF I'M BEING OBTUSE... it's not intentional, lol)

• Libra

[therealdb]

Vera 7.31 will only get tts. Maybe written commands, I’ve not tested it.

[LibraSun]

...and what does "MFA disabled" mean in this context?

[cw-kid]

@librasun presumably that's referring to your Amazon account and 2 factor authentication, needs to be disabled.

[LibraSun]

Ah, thanks for explaining. Could not figure out the meaning of "MFA" but now get it: "Multi-Factor Authentication"

[therealdb]

I have to update the doc, since I have MFA on my account, with Authenticator app and oath tool in place wonderfully working together. But this definitely requires openluup.

[therealdb]

I double checked and spoken commands are available even if jq is not installed:

-- execute a spoken command on Bedroom device
luup.call_action("urn:bochicchio-com:serviceId:VeraAlexa1",
   "RunCommand",
   {Command="-e textcommand:'Alexa, I’m leaving' -d 'Bedroom'"}, 666)

I've updated the doc as well.

[LibraSun]

Sweet!! Thanks. Now, if I could just get VOI to succeed in turning off the screen of my Echo Spot, which so far can only be done in person by voice ("Alexa, turn off the screen").

I've begged Amazon tech support to improve that feature to be responsive to automation, but so far no go.

[therealdb]

You should be able to do it via voice commands. I added a specific section in read me and I’ll probably create a specific service in next version.